eyalroz / cuda-kat

CUDA kernel author's tools
BSD 3-Clause "New" or "Revised" License
105 stars 8 forks source link

Add <algorithm> and <numeric> functions as thread-level primitives? #42

Open eyalroz opened 4 years ago

eyalroz commented 4 years ago

While it's rarely a great idea, for the sake of completeness, we may want to have implementations of the Add abstract <algorithm> and <numeric> algorithms which could be run by all threads without collaboration, each on its own data.

What do you think? Good idea? Bad idea?

See also issue #18.